LEAP Hand anthropomorphic robotic hand


Project Description:

The LEAP Hand is a low-cost, dexterous, and anthropomorphic robotic hand developed by Kenneth Shaw, Ananye Agrawal, and Deepak Pathak from Carnegie Mellon University. Designed for machine learning research, the LEAP Hand features a unique kinematic structure that ensures maximum dexterity in various finger poses. It’s cost-effective and can be assembled in four hours using readily available parts, totaling a cost of $2000. The hand can consistently exert large torques over extended periods and is capable of performing several manipulation tasks in real-world settings, such as visual teleoperation, learning from passive video data, and sim2real.


Resources and Documentation: